[19p-Z32-1~15] 23.1 Joint Session N "Informatics"

△ [19p-Z32-6] Maps for New Materials Search Generated by Machine Learning

〇Yuki Inada¹, Yukari Katsura^1,2, Haruhiko Morito³, Masaya Kumagai⁴, Kaoru Kimura¹ (1.Univ. of Tokyo, 2.NIMS, 3.Tohoku Univ., 4.Kyoto Univ.)

Keywords:Materials Informatics, Machine learning, New Materials Search

We made a machine learning prediction model which predict whether compounds can be synthesized or not from input elements combination. Probabilities of ternary compounds existence were predicted for 80x80x80 combinations of elements. These results were shown in 80 heatmaps. Based on these heatmaps, we carried out M-Al-Ge system compounds synthesis and compounds which have unknown XRD peaks were synthesized on 32 pattern elements combinations.